Data emerging as accelerator of AI, innovation, report says
By YIN MINGYUE and LIU BOQIAN in Guiyang | China Daily | Updated: 2026-08-29 06:52
China's data industry is expanding rapidly while new forms of data-related services are emerging as artificial intelligence becomes increasingly embedded in the economy, according to a report released on Friday at the China International Big Data Industry Expo 2026 in Guiyang, Guizhou province.
China's data industry reached 6.78 trillion yuan ($1.01 trillion) in 2025, up 15.7 percent year-on-year, according to the China Data Industry Development Report 2026 released by the National Data Administration's National Institute of Data Development.
Data-related software products and data-resource products involving token services together generated 4.99 trillion yuan, accounting for 73.6 percent of the industry's total output, it said.
The report also showed that the number of data enterprises nationwide reached 482,000 at the end of 2025, up 17.9 percent year-on-year, with financing activity particularly strong among companies engaged in large language models, synthetic simulation and embodied-intelligence services.
"The emergence of token services and other new forms of business is accelerating," said Hu Jianbo, director of the National Institute of Data Development.
The emergence of such new services is also changing how data supports the development of AI. Yang Jie, president of the Internet Society of China, said data serves as the foundation for AI, computing power provides the infrastructure, models act as the engine, and applications are the outlet, while tokens connect the different stages.
Wang Jingtao, deputy director of the Cyberspace Administration of China, said that as generative AI applications expand, tokens are increasingly showing their potential to be measured, priced and traded, providing a new way to observe the activity of AI and the depth of its industrial penetration.
As AI becomes more capable, the question is not only how data is measured and valued, but also what kind of data can make AI more capable. Zhou Bowen, director of the Shanghai Artificial Intelligence Laboratory, said AI is moving from solving tasks with clear answers toward more open-ended scientific and industrial innovation.
"In such settings, data is generated not only before a task begins, but throughout the process of experimentation, feedback and correction," Zhou said.
He described this emerging model as "environment as data", in which real-world environments provide models with feedback that can be used for further learning and improvement.
Such changes are creating new demands for data, computing power and real-world application scenarios. For Guizhou, a major data industry hub and the host province of the expo, the focus has increasingly shifted toward linking computing power, data resources and AI applications.
The province has opened more than 25,000 datasets to the public and developed 150 public-data products, while more than 170 AI application scenarios have been established in key sectors, according to Xu Lin, Party secretary of Guizhou. The province also has more than 60,000 data-related enterprises.
The changes are also being reflected in how companies organize and use digital resources.
Zhang Ping'an, chairman of Huawei Cloud, said token consumption in China has grown more than 1,000-fold over the past two years, while enterprise IT architecture is shifting from being "people-centered" toward being "agent-centered".
Contact the writers at yinmingyue@chinadaily.com.cn





















